Conventionally, a client and server type database system is offered. In the database system of this kind, a master database (hereinafter referred to as a maser DB) is provided in a server, a local database (hereinafter referred to as a local DB) is provided in a client, and data stored in the master DB is downloaded to the local DB of the client and desirable data processing is thus carried out in the client.
The data in the master DB is downloaded to the local DB of the client. Regardless of an open system to be operated in a web environment, therefore, the same data processing as that in connection to the server can be carried out in the client also in an offline state in which the client and the server are not connected through a network. In this case, however, it is necessary to guarantee integrity of the master DB and the local DB, and various mechanisms therefor are offered (for example, see Patent Documents 1 to 4).
Referring to the technique described in the Patent Document 1, there is provided means (a synchronous control section) for causing the data of the master DB and that of the local DB to be coincident with the newest contents. Also in the case in which addition, change, delete or the like of data occurs for the master DB of the server and the local DB of the client, consequently, it is possible to ensure integrity of the respective databases quickly and accurately.
Referring to the technique described in the Patent Document 2, the data stored in the master DB of the server is fetched and stored in the local DB of the client. The client executes specific processing by using local data, and carries out the same data processing over local data and master data corresponding to the local data. If the data processing is not normally ended, it is rolled back. Consequently, consistency of the master data and the local data is guaranteed.
Referring to the technique described in the Patent Document 3, every time the business application of the client refers to the data through the integrated database, the history of the data is registered as history information in the data history file. When the business application of the client is started, the history information is extracted from the data history file and is transmitted to the server. When the transmitted history information is different from the data version of the integrated database on the server side, data is extracted from the integrated database and is transferred to the client to update the integrated database on the client side. Consequently, useless data transfer or processing is omitted to increase a speed of processing.
Referring to the technique described in the Patent Document 4, when a retrieval request is made in the client, the client inquires the version of its own database to the server. When confirming that the version is the newest, the client retrieves its own database. On the other hand, when the client cannot confirm the newest version, it acquires the difference information about the database from the server and updates its own database and then carries out retrieval. Consequently, it is possible to suppress access to the server as greatly as possible.
Patent Document 1: Japanese Laid-Open Patent Publication No. 2001-142766
Patent Document 2: Japanese Laid-Open Patent Publication No. 2003-263356
Patent Document 3: Japanese Laid-Open Patent Publication No. 2000-330880
Patent Document 4: Japanese Laid-Open Patent Publication No. 2003-150595